Allow timewarp with Ion engines using a simplified physics simulation

[Grumman]

Ion engines as they exist right now cannot be used realistically because of the weaknesses of the simulation - since we have to watch the flight in real time during a burn, long slow burns are impractical. The reason for this, as we all know, is because our computer cannot necessarily cope with performing all the physics calculations the game requires 100,000 times faster than usual.

Fortunately, we don't need all those physics calculations.

Structural stresses: Ion engines produce very little thrust - a maximum of 2 kN (and even that can be reverted back to its previous lower value once we have the ability to perform realistic ion burns). Since we are only allowing time acceleration when that thrust is acting through the CoG we don't have to worry about centripetal force as it starts spinning out of control, and that force is not enough to break any part or join while not under timewarp, we can simply ignore structural stresses, just as we do in timewarp right now.

Changing Mass and CoG: Ion engines use very little fuel mass - a maximum of 1/20th of a kilogram per second. Simply ignoring this change in mass and CoG until you come back out of timewarp would do very little to change your flight path, and the effect of this approximation assuming your ship is slightly heavier than it is would only be a penalty so it cannot be exploited.

My Proposal:

- In addition to the current requirements to engage timewarp, allow timewarp if the only active engines are Ion engines AND the summed thrust vector of all those active Ion engines passes sufficiently close to the ship's current center of gravity.

- If this requirement is met and timewarp is used, calculate the ship's displacement over time during timewarp using two and only two acceleration vectors - the gravitation force applied by whatever celestial body's SoI the ship currently occupies, and a constant ion thrust vector based on the ship's mass, CoG and thrust at the time timewarp was entered.

- If the ship runs out of xenon or electricity, revert to non-timewarp as normal.

[tater]

Yeah, constant thrust trajectories would be cool. The big issue might not be time warp, but in calculating the spiral to show in map mode when planning a maneuver, however.

Say you want a low-thrust trajectory to Jool. Right now you drag a transfer, and do the burn. The specifics don't matter, it's X dv required. For a low-thrust trajectory, it's not the total dv that matters to show it, it's how that is applied over time. With the spacecraft mass changing, so it's not just the engines currently active. I suppose you could take total craft mass, and current active constant acceleration engines, and have it use a guess for dv added per second, and expect to have to do periodic corrections as the craft mass decreases (as you will no longer match the mapped trajectory).

[hobbsyoyo]

I'd support this fully.

However, I'd also like to see an SAS system that could actually handle the trajectory while on rails. I haven't unlocked ion engines in the new version of KSP but in the past, you had to stop the 4x time warp every minute or so because the SAS would wander off the chosen heading. This gets really frustrating and I think if you added in on-rails time warp, it would be much worse. Currently the SAS can't keep your vessel locked on a heading during on-rails timewarp even when you aren't thrusting. So I'd hate to go into 10000x time warp and instantly be way off-heading.

I'm not talking about trajectory corrections (i.e. the orbital path itself), just the fact that the SAS can't keep a steady lock on the chosen heading on the navbal. That needs to be fixed before we do on-rails ion drives in my opinion.

[MacroNova]

Did you know you can use physics time warp (like how you time warp in atmosphere) in space by holding alt? Very helpful with long ion burns.

[Grumman]

I'd support this fully.

However, I'd also like to see an SAS system that could actually handle the trajectory while on rails. I haven't unlocked ion engines in the new version of KSP but in the past, you had to stop the 4x time warp every minute or so because the SAS would wander off the chosen heading. This gets really frustrating and I think if you added in on-rails time warp, it would be much worse. Currently the SAS can't keep your vessel locked on a heading during on-rails timewarp even when you aren't thrusting. So I'd hate to go into 10000x time warp and instantly be way off-heading.

That's why I said it should only work after it has checked that the thrust vector can reasonably be approximated as having zero rotational component. Timewarp already negates all rotation, so this is a solved problem.

There's no reason why you couldn't build upon my proposed system at a later date, but this would at least provide a usable tool. The PB-X750 xenon tank contains three hours of fuel - with my proposal it would be possible to actually use it without spending at least 45 minutes doing nothing.

[NathanKell]

http://forum.kerbalspaceprogram.com/threads/70881-0-25-Orbit-Manipulator-Series

[NathanKell]

To be clear, I'm not answering in the spirit of "meh, a mod did it so it shouldn't be stock." I think it should be stock. I'm preemtively responding to anyone who says it's impossible.

[tater]

Does that mod actually create spiral orbits when you drag on a node?